Glioblastomas have markedly elevated activity of signal transduction pathways relative to low-grade gliomas. Analysis of human glioma samples using Western blot analysis with antibodies specific for the activated forms of these signaling components demonstrates that all of the GBMs show elevated levels of Erk. These data strongly indicate that the downstream readout of Ras in these tumors in vivo is through Raf, MEK, and Erk. Akt activity is elevated in about 60% of GBMs and is associated with phosphorylation of S6 kinase and 4EBP, both of which are downstream of mTOR. The elevation of activity of these proteins relative to normal brain or lower-grade gliomas is striking and alone suggests that these signaling pathways may be causally related to the formation of gliomas. The proof that the activity of these pathways can actually cause GBM is derived from mouse modeling studies with the RCAS/tv-a system where the combination of activated Ras and Akt results in the formation of GBMs in mice. Therefore, not only are these pathways elevated in human GBMs but also experimental elevation of these pathways in mice results in the formation of a very similar tumor. From these data we hypothesize that the signaling pathways are the etiology of at least some GBMs in humans. We propose to investigate three major specific questions that will be addressed in eight limited specific aims in this project. The first major question is whether the oncogenic effect of Ras in the formation of glioblastoma proceeds through Raf, Mek, and Erk. Specifically whether this pathway is necessary and sufficient for Ras activity in this context. The second major question is whether EGFR activity and PTEN loss functions through Akt and mTOR and whether mTOR activity is necessary and sufficient for the oncogenic effects of elevated Akt activity in glioblastomas. The critical importance of these first two questions relates to the available small molecule inhibitors that are currently in clinical trials. These drugs are specific for certain components within these signaling pathways but we do not know if these pathways are actually the oncogenic effectors for the formation of gliomas. In this grant we will address these questions using the RCAS/tv-a mouse modeling system for Ras + Akt-induced gliomas. [unreadable] [unreadable]